Go Down

Topic: Reading Pressure at different points of Wind Tunnel (Read 431 times) previous topic - next topic

abbmohpilot

Hello everyone!!

I'm a newcomer here and to Arduino in general :)

I'm working on improving a lab Wind Tunnel as part of a summer internship in my university.

As I have zero experience with arduinos. To keep it simple, I need to log the static pressure of 10 different points around the wind tunnel. Holes around the wind tunnel would be connected to pressure gauges with plastic tubes.


My questions are:

- Is it possible to use only 1 arduino card (whichever is best) to log 10 different pressure points?

- What kind of pressure sensor should I use too? (It should read a range of 2 bars)

Many thanks in advance! 0:)

Robin2

My questions are:

- Is it possible to use only 1 arduino card (whichever is best) to log 10 different pressure points?

- What kind of pressure sensor should I use too? (It should read a range of 2 bars)
I suspect you know more about your second question (wind tunnel pressure sensors) than anyone here - or your project colleagues will.

And the answer to your second question is required before we can answer the first question.

Post a link to the datasheet for the sensors.

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

dave-in-nj

#2
Jul 23, 2018, 12:35 pm Last Edit: Jul 23, 2018, 12:52 pm by dave-in-nj
there are many ways to read 10 analog points, but that is the second question.

the first is what sensor to use.
once you know that, the inputs become simple.

to read 2 bars, your sensor full scale needs to be 2 bars or more.
I would consider putting all the sensors in one place and run tubing from the point back to a central panel.

Back in the old days of pneumatic sensors, we would run 250 feet without a problem.  usually 8mm or larger for that distance.


the proper sensor is one that has a full scale of 2 bars, but the hobby way is to use a cheap sensor.  instead of paying $100 for each sensor, you could use a BMP280

use SPI (serial reading of sensors) to read the pressure of each sensor.

I would use 6mm tubing and get multiple fish tank valves

air valves

then get single valves and tee fittings.
the sensor goes to the tee, the one line to the manifold, the other to a single valve, then to the tubing that goes to the point in the tunnel you want to measure.

that way, you can easily connect all of them to a single point to calibrate.  the manifold should be much larger dia.
the manifold should have two points, the highest pressure you measure, and open to the room air.
a third point in the middle will allow an additional reference point.

anwofis

Couldn't you use an anemometer and calculate the pressure from the wind speed?

dave-in-nj

Couldn't you use an anemometer and calculate the pressure from the wind speed?
put an anemometer in a ballon.  
run test for pressure
validate hypotheses.


Robin2

From my attempts (many years ago) to understand aerodynamics I believe one needs a very sensitive pressure sensor for a wind tunnel.

And a wind tunnel is an expensive thing so it would not make any sense to use poor quality sensors.

...R
Two or three hours spent thinking and reading documentation solves most programming problems.

dave-in-nj

From my attempts (many years ago) to understand aerodynamics I believe one needs a very sensitive pressure sensor for a wind tunnel.

And a wind tunnel is an expensive thing so it would not make any sense to use poor quality sensors.

...R
I absolutly agree.  
having been involved in air flow and room pressuirzation for clean rooms, I doubt that the project will allow for $1,000 sensors.
however, there is a company that uses one sensor and uses a multi-port valve to test each pressure.
much more cost effective when doing multiple dozens of pressures along a wing.   Scan-valve or some such..... 
However, a simple installation of very cheap parts and offer some interesting insights.
As a note, the flows, eddy currents, stagnation, boundary layers, etc will all cause effects.
without getting into Reynolds Numbers, the general idea would be to just get a reading.
I am assuming just fan and tunnel performance and not subject testing.
UNRELATED :
As a note, the scan-valve concept is something I was thinking about for plant watering.
I have mentioned in the past,  pump water up to a quart container.
then rotate/move the desires pipe under the nozzle, then open and let it drain into the corresponding plant.
alas, needs a soil sensor for each plant. 
one pump, one valve, and one servo, water dozen plants individually.

DrDiettrich

Dave seems to be familiar with practical air pressure measurement :-)

Which frequencies do you expect in the static pressure signals, which of these are of interest (signal bandwidth)?
Shall pressure measured only at the walls, or also across the channel?

In addition to the tubing some more signal smoothing may be required, either mechanically or by low pass filters on the analog signals. Please note that the orientation of the holes or valves in the channel walls can affect the readings seriously.

abbmohpilot

First, Thank you all very much for your replies!

So to answer back some of your questions:

- I still have to find the appropriate sensor to use

then get single valves and tee fittings.
the sensor goes to the tee, the one line to the manifold, the other to a single valve, then to the tubing that goes to the point in the tunnel you want to measure.

that way, you can easily connect all of them to a single point to calibrate.  the manifold should be much larger dia.
the manifold should have two points, the highest pressure you measure, and open to the room air.
a third point in the middle will allow an additional reference point.
However, a simple installation of very cheap parts and offer some interesting insights.
As a note, the flows, eddy currents, stagnation, boundary layers, etc will all cause effects.
without getting into Reynolds Numbers, the general idea would be to just get a reading.
I am assuming just fan and tunnel performance and not subject testing.

- It's true we don't really need to measure all 10 points all the time, so such a system is a very good idea :D

Which frequencies do you expect in the static pressure signals, which of these are of interest (signal bandwidth)?
Shall pressure measured only at the walls, or also across the channel?
- Currently we just need to measure the pressure at the walls, with holes perpendicular to them :)


So for the sensor, if I can use only one, would this one be suited?


http://www.te.com/commerce/DocumentDelivery/DDEController?Action=showdoc&DocId=Data+Sheet%7FMS4525DO%7FB2%7Fpdf%7FEnglish%7FENG_DS_MS4525DO_B2.pdf

TomGeorge

#9
Jul 24, 2018, 10:54 am Last Edit: Jul 24, 2018, 10:55 am by TomGeorge
Hi,
Welcome to the forum.

I think some have not quite read your first post, you have 10 points already setup to measure static pressure.
You need a pressure sensor that will read up to 2bar that will interface with hopefully an Arduino.

Am I right?

The Arduino controllers can interface in many ways, analog volts (0-5V), I2C and SPI comms without any major external circuitry.
However if you have 10 sensors  you will need an input that can cater for them, I2C or SPI, or you use 10 analog inputs,

The sensor you have linked to can be SPI or I2C, so they would make the grade.

OPs, sensor;


How often do you need to read the sensors and how do you aim to display or log the data.
The Arduino controllers do  have a serial connection to the PC to send data, or you could add a SD card to store your data.

Can you please tell us your electronics, programming, Arduino, hardware experience?

Tom.. :)

Everything runs on smoke, let the smoke out, it stops running....

abbmohpilot

#10
Jul 24, 2018, 11:45 am Last Edit: Jul 24, 2018, 11:46 am by abbmohpilot
Hello Tom! :)

Correct I have to measure static at pressure at 10 different points around a wind tunnel.

Frequency isn't much of a problem, even having one value per second is enough as values will be read for each speed setting of the wind tunnel.

I'd like to be able to visualize the data on a a histogram, showing each tapping point in X-axis and the pressure in Y-axis.

I have very basic Arduino knowledge, the most advanced thing I've done is reading the velocity off a PIXHAWK sensor (https://goo.gl/897hAQ) connected to an Arduino Nano.

It is also what inspired me to use an Arduino based sensor reading solution, instead of a multi tube manometer traditionally used in aerodynamics (pictured below)


Billhigdon

I have 2 friends who have a wind tunnel & instrumented it quite well. I'll get you some pictures of the instrumentation setup, after they return from Oshkosh next week. If your using a combination Pitot Static tube you'll want to use a differential sensor for each pickup.

Jregester

I am quite familiar with high accuracy measurements. To get you going in the right direction more information is needed.

First would be what accuracy is needed.

Second would be if you're using gauge pressure or dP sensors. (I would assume dP otherwise relative error will be difficult to solve)

Third would be range parameters. You may be able to measure to 1% accuracy, but if dP between point A and point B is only 5%, your 1% measurement only gives you +/- 20%

Forth, do you need any certifications for your hardware?

For university level stuff I would guess 12 bit AtoD accuracy would be a minimum and may need 16 bit once you crunch the numbers.

Setra makes good quality pressure sensors. Might want to call one of their tech sales people and let them do a product selection for you. Normal industry outputs are 0-10v or 4-20mA but you might be able to specify if needed.

As much as I like Arduinos, I am thinking that if you have the budget, National Instruments would be a better fit. Lab view is pretty powerful and made for these types of applications.

dave-in-nj

I noticed some people posting about velocity pressure as a measurement. but the OP posted about static pressure, not looking for flow.

a dp sensor for velocity pressure would have a very small range, whereas a static pressure sensor would need to be greater than 2 bar.

the 'sensor' needs to be designed and manufactured for static pressure.   typically a well conformed static pressure penetration in the side wall, or on a static only tube  that is similar to a Pitot tube  (concentric holes around the axis, sharing the tube as a common port, whose areas, combined, are less than 1/10th of the area of the tube diameter.

Setra certainly is a good name and have a wide variety of offerings. 

As for selection, I would highly recommend you do a bit of research of what others doing the exact same thing, on wind tunnels have been doing.

Dywer is a supplier of devices, but a careful review of data sheets is needed to make sure the device is of the quality required.

there are wind tunnel instrumentation suppliers.  Some time with Google will help.

as for connecting the sensors, the world is open.  cost of the bits and your budget will determine what works best. 
accuracy costs money, how accurate can you afford to be ?

at 2 bar, you are well into the PSI ranges so some smaller sensors might allow you to play without a large budget, but also without any traceability on your instruments.



Billhigdon

abbmohpilot
I just sent you a PM with some info

Go Up